Number 1 Volume 18 January 2012 Journal of Engineering

DOUBLE DIFFUSIVE FREE CONVECTION IN A PACKED BED

SQUARE ENCLOSURE BY USING LOCAL THERMAL
NON-EQUILIBRIUM (LTNE) MODEL
Ahmed N. Mehdy
Assistant Lecturer
Department of Mechanical
Engineering
Engineering college

ABSTRACT
In the present study, free convection heat and mass transfer of fluid in a square packed bed enclosure is
numerically investigated. For the considered geometrical shape, the left vertical wall of enclosure was
assumed to be kept at high temperature and concentration while the opposite wall was kept at low
temperature and concentration with insulating both the top and bottom walls of enclosure. The Brinkman–
Forchheimer extended Darcy model was used to solve the momentum equations, while the energy equations
for fluid and solid phases were solved by using the local thermal non-equilibrium (LTNE) model.
Computations are performed for a range of the Darcy number from 10-5 to 10-1, the porosity from 0.5 to 0.9,
and buoyancy ratio from -15 to 15. The results showed that both the buoyancy ratio and the packed bed
characteristics have significant effect on each one of the flow field, heat transfer and mass transfer.

‫اﻟﺨﻼﺻﺔ‬
‫ ﻟﻠﻨﻤ ﻮذج اﻟﻬﻨﺪﺳ ﻲ‬.‫ اﻧﺘﻘﺎل اﻟﺤﺮارة واﻟﻜﺘﻠﺔ ﺑﺎﻟﺤﻤ ﻞ اﻟﺤ ﺮ ﺧ ﻼل ﻃﺒﻘ ﺎت ﻣﺤ ﺸﻮة ﻓ ﻲ ﺗﺠﻮﻳ ﻒ ﻣﺮﺑ ﻊ ﺗ ﻢ دراﺳ ﺘﻪ ﻋ ﺪدﻳًﺎ‬،‫ﺧﻼل اﻟﺪراﺳﺔ اﻟﺤﺎﻟﻴﺔ‬
‫ ﺗﻢ اﻋﺘﺒﺎر اﻟﺠﺪار اﻟﻌﻤ ﻮدي اﻷﻳ ﺴﺮ ﻟﻠﺘﺠﻮﻳ ﻒ ﻣﺤﻔ ﻮظ ﻋﻨ ﺪ درﺟ ﺔ ﺣ ﺮارة وﺗﺮآﻴ ﺰ ﻣ ﺮﺗﻔﻌﻴﻦ ﺑﻴﻨﻤ ﺎ اﻟﺠ ﺪار اﻟﻤﻘﺎﺑ ﻞ ﻗ ﺪ ﺣُﻔ ﻆ ﻋﻨ ﺪ درﺟ ﺔ‬،‫اﻟﻤﻔﺘﺮض‬
‫ ﻓﻮرﺷﻤﺎﻳﺮ اﻟﻤﺸﺘﻖ ﻣ ﻦ ﻧﻤ ﻮذج دارﺳ ﻲ ﻓ ﻲ ﺣ ﻞ‬-‫ ﺗﻢ اﺳﺘﺨﺪام ﻧﻤﻮذج ﺑﺮﻧﻜﻤﺎن‬.‫ﺣﺮارة وﺗﺮآﻴﺰ واﻃﺌﻴﻦ ﻣﻊ ﻋﺰل آﻞ ﻣﻦ اﻟﺠﺪارﻳﻦ اﻟﻌﻠﻮي واﻟﺴﻔﻠﻲ‬
‫ ﻟﻘ ﺪ‬.(LTNE) ‫ ﺑﻴﻨﻤﺎ ﺗﻢ ﺣﻞ ﻣﻌﺎدﻻت اﻟﻄﺎﻗﺔ ﻟﻜﻞ ﻣﻦ اﻟﻄﻮرﻳﻦ اﻟﻤﺎﺋﻊ واﻟﺼﻠﺐ ﺑﺎﺳﺘﺨﺪام ﻧﻤﻮذج ﻋﺪم اﻻﺗﺰان اﻟﺤﺮاري اﻟﻤ ﻮﻗﻌﻲ‬،‫ﻣﻌﺎدﻻت اﻟﺰﺧﻢ‬
ً ‫ أﻇﻬ ﺮت اﻟﻨﺘ ﺎﺋﺞ ﺑ ﺄن آ‬،15 ‫ إﻟ ﻰ‬-15 ‫ وﻧ ﺴﺒﺔ اﻟﻄﻔ ﻮ ﻣ ﻦ‬،0.9 ‫ إﻟ ﻰ‬0.5 ‫ اﻟﻤﺴﺎﻣﻴﺔ ﻣﻦ‬،10-1 ‫ إﻟﻰ‬10-5 ‫أﻧﺠﺰت اﻟﺤﺴﺎﺑﺎت ﻟﻤﺪى رﻗﻢ دارﺳﻲ ﻣﻦ‬
‫ﻼ‬
.‫ اﻧﺘﻘﺎل اﻟﺤﺮارة واﻧﺘﻘﺎل اﻟﻜﺘﻠﺔ‬،‫ﻣﻦ ﻧﺴﺒﺔ اﻟﻄﻔﻮ وﺧﺼﺎﺋﺺ اﻟﻄﺒﻘﺎت اﻟﻤﺤﺸﻮة ﺗﻤﺘﻠﻜﺎن ﺗﺄﺛﻴﺮًا ﻣﻬﻤًﺎ ﻋﻠﻰ آﻞ ﻣﻦ ﺣﻘﻞ اﻟﺠﺮﻳﺎن‬

Keywords: Double Diffusive, Free Convection, Packed Bed, Local thermal non-equilibrium.

121
Ahmed N. Mehdy DOUBLE DIFFUSIVE FREE CONVECTION IN A PACKED
BED SQUARE ENCLOSURE BY USING LOCAL THERMAL
NON-EQUILIBRIUM (LTNE) MODEL

INTRODUCTION buoyancy forces is necessary in order to control

Double-diffusive convection or thermosolutal these processes. (Chaudhary and Jain, 2007)
convection is generally referred to a fluid flow studied the MHD flow past an infinite vertical
generated by buoyancy effects due to both oscillating plate through porous medium with the
temperature and solutal concentration gradients. presence of free convection and mass transfer
This type of flow is encountered in natural and analytically by using Laplace-transform
technological applications. Such applications technique. (Mohamed, 2009) analyzed the
include the growth of crystals, solar energy double-diffusive convection-radiation interaction
systems, welding processes, thermal insulations. for the unsteady MHD flow over a semi-infinite
The term double diffusive-convection is now vertical moving porous plate embedded in a
widely accepted for all processes which involve porous medium in the presence of thermal &
simultaneous thermal and solutal concentration solutal diffusion and heat generation. A numerical
gradients and provides an explanation for a study of the unsteady free convection and mass
number of natural phenomena. Because of the transfer flow of an electrically conducting fluid
coupling between the fluid velocity field and the past an infinite vertical porous plate in the
diffusive (thermal and solutal concentration) presence of a transverse magnetic field was
fields, double-diffusive convection is more presented by (Shariful et. al. , 2005). (Bukhari,
complex than the convection flow which is 2003) applied a linear stability analysis, using the
associated with a single diffusive scalar, and many spectral Chebyshev polynomial method, to obtain
different behaviours may be expected. Such numerically the solution of a multi-layer system
double-diffusive processes occur in many fields, consisting of the finger convection onset in a fluid
including chemical engineering (drying, cleaning layer overlying a porous layer. (Saha and
operations, evaporations, condensation, Hossain, 2004) studied numerically the laminar
sublimation, deposition of thin films, energy doubly diffusive free convection flows along an
storage in solar ponds, roll-over in storage tanks isothermal vertical finite plate immersed in a
containing liquefied natural gas, solution mining stable thermally stratified fluid by using an
of salt caverns for crude oil storage, casting of implicit finite difference method and local non-
metal alloys and photosynthesis), oceanography similarity method. (Hajri et. al. , 2007) presented
(melting and cooling near ice surfaces, sea water a numerical simulation for the steady double-
intrusion into freshwater lakes and the formation diffusive natural convection in a triangular cavity
of layered or columnar structures during by using equal finite elements method. A
crystallisation of igneous intrusions in earth's numerical study was presented from (Mamou et.
crust), geophysics (dispersion of dissolvent al. , 2001) for the unsteady double-diffusive
materials or particulate matter in flows), etc. A convection in a two-dimensional horizontal
clear understanding of the nature of the interaction confined enclosure by using the finite element
between thermal and mass or solutal concentration technique. (Masuda et. al. , 2002) presented a
numerical simulation by using finite differences

122
 Number 1 Volume 18 January 2012 Journal of Engineering

method for the unsteady two-dimensional double- MATHEMATICAL MODEL

diffusive convection in porous rectangular Geometrical Shape of Studied Problem
enclosure. After that, (Masuda et. al. , 2008)s The schematic view of the studied problem is
studied the peculiar oscillating convection which shown in Fig.1. The square enclosure has a side
is observed when two-dimensional double- length (a) and it is filled with a saturated packed
diffusive convection in porous medium is bed. The left vertical wall of the square enclosure
analyzed numerically. (Khanafer and Vafai, is kept at high temperature and concentration
2002) presented a numerical study of mixed- while the opposite wall is kept at low temperature
convection heat and mass transport in a lid-driven and concentration, finally, the horizontal walls are
square enclosure filled with a non-Darcian fluid- insulated.
saturated porous medium by using the finite
volumes technique. (Rahli and Bouhadef, 2004) Governing Equations
studied numerically the double-diffusive natural In the model development, the following
convection in a partially porous square enclosure assumptions are adopted; the working fluid has a
with the presence of inclination effect by using the Prandtl number Pr=0.71 and assumed to be
control volume method and the power law incompressible and Newtonian fluid, no phase
scheme. (Wang et. al., 2007) presented a change occurs and the process is in a steady state,
numerical investigation of natural convection of the thermo physical properties of the fluid are
fluid (without mass transfer) in an inclined square assumed to be constant except the density
enclosure filled with porous medium and variation in the buoyancy force, which is
submitted to a strong magnetic field by using approximated according to the Boussinesq
(LTNE) model. However, the buoyancy force due approximation. This variation, due to both
to the double-diffusive effect has received more temperature and solutal concentration gradients
attention in the literature. But all the previous can be described as follow (Khanafer and Vafai,
studies which deal with the double-diffusive 2002);
convection in porous mediums assume that the ρ = ρ0 [1− βT (T −TL ) − βC (c −cL )] (1)
porous medium in thermo dynamical equilibrium Where βT and βC are the coefficients of the
with the fluid which flow inside it, except (Wang thermal and solutal expansions, which are defined
et. al., 2007) which studied the pure convection as follow (Khanafer and Vafai, 2002);
without mass transfer in the presence of heat
1 ⎛ ∂ρ ⎞
transfer between the fluid and the porous media. βT = − ⎜ ⎟ &
ρ0 ⎝ ∂T ⎠p,c
The present study gives more attention to the (2)
1 ⎛ ∂ρ ⎞
interaction between the effect of the double- βC = − ⎜ ⎟
ρ0 ⎝ ∂c ⎠p,T
diffusive convection and the effect of heat transfer
The Brinkman–Forchheimer extended Darcy
between the porous medium and the fluid which
model is used to solve the momentum equations
flows inside it.
while the energy equations for fluid and solid
phases are solved with the local thermal non-
123
Ahmed N. Mehdy DOUBLE DIFFUSIVE FREE CONVECTION IN A PACKED
BED SQUARE ENCLOSURE BY USING LOCAL THERMAL
NON-EQUILIBRIUM (LTNE) MODEL
equilibrium (LTNE) model. Thus, the governing 1 .75
F = (9 .a )
equations for the present study will take the 150 ε 3
following forms as in (Amiri and Vafai, 1998) 6 (1 − ε )
a sf = (9 .b )
and (Khanafer and Vafai, 2002); dp
- Continuity equation ⎡ ⎛ρ u d ⎞ ⎤
0 .6

⎢ ⎜ f p ⎟ ⎥
∂u ∂v h sf = k f ⎢ 2 + 1 .1 Pr 1 3 ⎜ ⎟ ⎥ (9 .c )
+ =0 (3) ⎜ µf ⎟ ⎥
∂x ∂y ⎢ ⎝ ⎠ ⎦
⎣
- Momentum equations
ρf ⎛ ∂u ∂u ⎞ where the sphere particle diameter dp can be
⎜⎜u + v ⎟⎟ =
ε2 ⎝ ∂x ∂y ⎠ computed as follow (Amiri and Vafai, 1998);
(4)
∂p µf ⎛ ∂2u ∂2u ⎞ µf F ρf u u 150 K
− + ⎜⎜ 2 + 2 ⎟⎟ − u+ d = (1 − ε ) ( 9 .d )
∂x ε ε 3
p
⎝ ∂x ∂y ⎠ K K

while the effective thermal conductivity kfeff and

kseff in fluid and solid phase energy equations and
the mean thermal diffusivity αm can be computed
as follow (Wang et. al. , 2007);
k feff = ε k f ( 10 . a )
(5)
k seff = (1 − ε ) k s ( 10 . b )
- Fluid phase energy equation k + k
α =
feff seff
( 10 . c )
ρ
m
∂T f ∂T f ⎞ cp
⎛ f f
ρ f cp f ⎜⎜ u +v ⎟=
⎝ ∂x ∂ y ⎟⎠
Now we introduce the following non-
⎛ ∂ 2T f ∂ 2T f ⎞
k feff ⎜ + ⎟ + a sf h sf (T s − T f ) dimensional quantities and parameters as in
⎜ ∂x 2 ∂ y 2 ⎟
⎝ ⎠ (Khanafer and Vafai, 2002) and (Wang et. al. ,
(6)
2007);
- Solid phase energy equation

⎛ ∂ 2Ts ∂ 2Ts ⎞
k seff ⎜⎜ + ⎟⎟ + a sf h sf (T f − T s ) = 0
⎝ ∂x ∂y 2
2
⎠
(7)
- Solutal concentration equation

∂c ∂c ⎛ ∂ 2c ∂ 2c ⎞
u +v = D ⎜⎜ 2 + ⎟
2 ⎟ (8)
∂x ∂y ⎝ ∂ x ∂ y ⎠
The geometric function F, specific surface area
of the packed bed asf and the fluid-to-solid heat
transfer coefficient in a packed bed hsf are
determined as suggested by (Amiri and Vafai,
1998);

124
 Number 1 Volume 18 January 2012 Journal of Engineering

x y ⎫ 1 ⎛ ∂V ∂V ⎞
X = , Y = 2 ⎜
U +V ⎟=
a a ⎪ ε ⎝ ∂X ∂Y ⎠
⎪
au av ⎪ ∂ P Pr ⎛ ∂ 2V ∂ 2V ⎞
U = , V = , − + ⎜⎜ + ⎟⎟ +
αm αm ⎪ ∂Y ε ⎝ ∂X
2
∂Y 2 ⎠
⎪
p a2 ⎪ FUV
Pr 2 GrT (θ f + N C )−
P= Pr
ρ f αm 2 ⎪ V − (14 )
⎪ Da Da
⎬ (11 .a )
(T f − T L ) ⎪
θf = ,
(T H − T L ) ⎪
(T s − T L )
⎪ (1 + Λ )⎛⎜⎜ U ∂∂θX
−1 f
+V
∂θ f ⎞
⎟=
∂ Y ⎟⎠
θs = , ⎪ ⎝
(T H − T L ) ⎪
⎪ ⎛∂ θf2
∂ 2θ f ⎞
⎜ + ⎟ + ξ (θ s − θ f )
(c − c L ) ⎪ ⎜ ∂X 2 ∂Y 2 ⎟
C= ⎪ ⎝ ⎠
(c H − c L ) ⎭
(15)
µ f cp f νf ⎫
Pr = = ⎪ ⎛ ∂ 2θ s ∂ 2θ s ⎞
k feff + k seff αm ⎪ 0 = ⎜⎜ + ⎟⎟ + Λ ξ (θ f − θ s ) (16)
⎝ ∂X ∂Y 2
2
⎠
g a β T (T H − T L ) ⎪
3
, GrT = ⎪
ν 2f ⎪ ∂C ∂C Pr ⎛ ∂ 2 C ∂ 2 C ⎞
U +V = ⎜ + ⎟ (17)
⎪ ∂X ∂Y Sc ⎜⎝ ∂X 2 ∂Y 2 ⎟⎠
g a 3 β C (c H − c L )
GrC = , ⎪
νf2 ⎪⎪ By using the (stream function-vorticity)
⎬ (11 .b )
νf formulation we will reduce the dependent
K ⎪
Da = 2 , Sc = ⎪ variables to only five variables by differentiating
a D ⎪ eq.(13) with respect to (Y) and differentiating
β C ( c H − c L ) GrC ⎪
N = = , ⎪ eq.(14) with respect to (X), after that we subtract
β T (T H − T L ) GrT
⎪ the first of the two resulted equations from the
k feff a sf h sf a 2 ⎪ second to eliminate the pressure terms from the
Λ= , ξ = ⎪
k seff k feff ⎪⎭ momentum equations, thus, eqs.(12, 13 & 14) will
be transformed to the following equations;
By substituting eqs.(11.a & 11.b) in eqs.(3, 4,
∂ 2ψ ∂ 2ψ
5, 6, 7 & 8), we get the dimensionless forms of + = −ω (17)
∂X 2 ∂Y 2
governing equations as follow;
∂ω ∂ω
∂U ∂V U +V =
+ =0 (12) ∂X ∂Y
∂X ∂Y ⎛ ∂2ω ∂2ω⎞
ε Pr⎜⎜ 2 + 2 ⎟⎟ (18)
1 ⎛ ∂U ∂U ⎞ ⎝ ∂X ∂Y ⎠
2 ⎜
U +V ⎟=
ε ⎝ ∂X ∂Y ⎠ ⎡ 2 ⎛ ∂θf ⎤
∂C ⎞
∂ P Pr ⎛ ∂ U ∂ U
2 2
⎞ ⎢Pr GrT ⎜⎜ + N ⎟⎟ − ⎥
− + ⎜ + ⎟⎟ − (13) ⎢ ⎝ ∂X ∂X ⎠ ⎥
∂X ε ⎜⎝ ∂ X 2 ∂ Y 2 ⎠ +ε2 ⎢⎛ ⎞ ⎛ ⎥
⎢⎜ Pr F U ⎟ F ⎜ ∂U ∂U ⎞⎟⎥
Pr FUU ⎢⎜ + ⎟ω − ⎜V −U ⎟⎥
U − ⎢⎜ Da Da ⎟ Da⎜ ∂X ∂Y ⎟⎥
Da Da ⎣⎝ ⎠ ⎝ ⎠⎦

125
Ahmed N. Mehdy DOUBLE DIFFUSIVE FREE CONVECTION IN A PACKED
BED SQUARE ENCLOSURE BY USING LOCAL THERMAL
NON-EQUILIBRIUM (LTNE) MODEL
where ψ and ω are the (stream function & NUMERICAL SOLUTION
The governing equations for ψ, ω, θf, θs & C
vorticity) respectively, and they are defined as
can be written in a common form for the
follow;
(convection-diffusion) problem as follow
∂ψ ∂ψ
U = & V = − (19 .a ) (Versteeg and Malalasekera, 1995);
∂Y ∂X
∂V ∂U ∂
ω = − (19 .b ) (ρ u iΦ )=
∂X ∂Y ∂X i

After getting the final values of all dependent ∂ ⎛ ∂Φ ⎞

⎜⎜ Γ ⎟⎟ + S ( 22 )
variables in the flow field, calculations will be ∂X i ⎝ ∂X i ⎠
made for local and mean Nusselt and Sherwood
where the general scalar Φ stands for any one of
numbers, where the local Nusselt and Sherwood
the dependent variables under consideration, the
numbers at the hot wall can be found as in (Wang
diffusion coeffecient Γ and the source term S in
et. al. , 2007) and (Khanafer and Vafai, 2002);
the cartesian form are listed below for each
⎛ ∂θ f ∂θ ⎞
Nul = ε ⎜ + Λ−1 s ⎟ (20.a) governing equation;
⎜ ∂X ∂X ⎟
⎝ X =0 X =0 ⎠ - Stream function equation
∂C Φ =ψ , Γ = 1 , S = ω ( 23 .a )
Sh = ( 20 . b )
∂X
l
X = 0
- Vorticity equation
Boundary Conditions Φ = ω , Γ = ε Pr , S =
The hydrodynamic boundary conditions for the ⎡ 2 ⎛ ∂θ f ∂C ⎞
⎢Pr GrT ⎜⎜ +N ⎟−
present problem at all enclosure walls will obey to ⎢ ⎝ ∂X ∂X ⎟⎠
the non-slip condition, while the thermal and ε 2 ⎢⎛
⎢⎜ Pr F U ⎞⎟
solutal boundary conditions are (the left side wall ⎢⎜ + ⎟ω
⎢⎜ Da Da ⎟
was kept at high temperature and solutal ⎣⎝ ⎠
concentration, the right side wall was kept at low ⎛ ⎞⎤
F ⎜ ∂U ∂U ⎟⎥
temperature and solutal concentration and finally − ⎜ V − U ⎟⎥ (23.b)
Da ⎜ ∂X ∂Y ⎟⎥
each one of the top and bottom walls were kept ⎝ ⎠⎦
insulated), thus the boundary conditions will be as
- Fluid phase energy equation
follow;
1 ξ (θ s − θ f )
U = 0, V = 0, θ f = θs = C = 1 ⎫ Φ =θ f , Γ = −1
,S = ( 23 .c )
⎪ 1+ Λ 1 + Λ−1
at X = 0 ; ⎪
U = 0, V = 0, θ f = θs = C = 0 ⎪ - Solid phase energy equation
⎪
at X = 1; ⎬ (21) Φ = θ s , Γ = 1 , S = Λ ξ (θ f − θ s ) ( 23 .d )
⎪
∂θ f ∂θ s ∂C
U = 0, V = 0, = = =0 ⎪
∂Y ∂Y ∂Y ⎪ - Solutal concentration equation
⎪
at Y = 0 & 1 ⎭ Pr
Φ =C , Γ = , S =0 ( 23 .e )
Sc

126
 Number 1 Volume 18 January 2012 Journal of Engineering

The numerical solution of the governing - Solid phase energy equation

equations will be made according to the finite S u = Λ ξ θ f ⎫⎪
volume method to transform the governing ⎬ ( 26 . d )
S p = −Λ ξ ⎪⎭
equations from partial differential form to discrete
- Solutal concentration equation
algebraic form, this method is based on principle
S = 0 ⎫
of dividing the flow field to a number of volume u
⎬ ( 26 .e )
S = 0 ⎭
elements, each one of them is called (control p

volume), after that a discretization process A computational program was written in

(Versteeg and Malalasekera, 1995) was carried Fortran-90 language to compute the values of the
out by integrating eq.(22) (of the general required variables, The discretized algebraic
conservation) over a control volume element, equations are solved by the tri-diagonal matrix
where this equation will be as follow; algorithm (TDMA). The used mesh size is
aP ΦP = aE ΦE + aW ΦW + aN ΦN + aS ΦS + Su (24) (60×60). Relaxation factors of about (0.7–0.9) are
used for all dependent variables, Convergence
where; was measured in terms of the maximum change in
a P = a E + aW + a N + a S − S P ( 25 ) each variable during an iteration where the
maximum change allowed for convergence check
the source coefficients Su and Sp represent the was 10-6.
source terms of the discrete equation and their
values for each governing equation are listed as RESULTS AND DISCUSSION
follow; All solution were carried out for solution of
- For the stream function equation (Pr=0.71 & Sc=0.25) at GrT= 105, where the
S u = ω ⎫ numerical code which is used in the present
⎬ ( 26 . a )
S = 0 ⎭ investigation has been carried out for number of
p
simulations for a wide range of controlling
- For the vorticity equation
parameters such as buoyancy ratio, Darcy number,
⎡ 2 ⎛ ∂θ f ∂C ⎞ ⎤ ⎫
⎢Pr GrT ⎜⎜ + N ⎟⎟ −⎥ ⎪ and porosity of the packed bed. Figs.(2 - 9) show
⎢ ⎝ ∂X ∂X ⎠ ⎥ ⎪ the effect of buoyancy ratio for different values of
Su = ε ⎢
2
⎥ ⎪
⎛ ∂ U ⎞⎟ ⎥
⎢ F ⎜ ∂U ⎪ (-15≤N≤15) on the stream function, fluid phase
⎢ ⎜V −U ⎥ ⎪
⎢ Da ⎜⎝ ∂X ∂Y ⎟⎟ ⎥ ⎬ (26.b) temperature and solutal concentration contours
⎣ ⎠⎦ ⎪ respectively at Da=10-1 and ε=0.9. Figs.[(2.a),
⎪
⎛ ⎞ ⎪
⎜ Pr F U ⎟ (2.b) & (2.c)] represent these contours for the case
S p = −ε ⎜ + 2
⎟ ⎪
⎜ Da Da ⎟ ⎪ of single diffusing effect at N=0, where there is
⎝ ⎠ ⎭
only thermal diffusing without solutal diffusing. It
is clear from the stream function contour at N=0
- Fluid phase energy equation
and as a result of the thermal buoyancy effect, the
ξ θ ⎫
S = s
−1 ⎪⎪ fluid at the left-hand side hot wall will be lighter
1 + Λ
u

⎬ ( 26 . c ) than in other locations while the fluid at the right-

ξ ⎪
S = − −1 ⎪⎭
p
1 + Λ hand side cold wall will be heavier than in other
locations, so, the fluid particles move upward
127
Ahmed N. Mehdy DOUBLE DIFFUSIVE FREE CONVECTION IN A PACKED
BED SQUARE ENCLOSURE BY USING LOCAL THERMAL
NON-EQUILIBRIUM (LTNE) MODEL
along the hot wall while they move downward fluid density will increase with increase of the
along the cold wall, and thus, the flow will take solutal concentration, and as a result, the flow
the direction of clockwise. The fluid phase begins to reverse its direction, So, the flow case at
temperature contour at N=0 indicates that the N=-1 as it is shown in Figs.[(6.a), (6.b) & (6.c)]
temperature levels will decrease gradually from represents a conversion point in the flow direction
the hot wall towards the cold wall, where the hot from clockwise to counterclockwise as it is shown
fluid rises up along the left-hand side hot wall and from the stream function distribution at N=-1,
descends along the right-hand side cold wall where the clockwise main central vortex which
because of the thermal buoyancy effect. Finally formed at each of the previous cases of positive
the solutal concentration contour at N=0 indicates buoyancy ratio will divide into two main vortices
that the solutal concentration will be maximum at in two opposite directions, also the distribution of
the hot regions and decreases gradually with the both the fluid phase temperature and the solutal
decrease of the fluid phase temperature. After concentration at N=-1 will begin to change their
that, and with increasing the positive buoyancy directions and their gradient will be at the
ratio N>0 as shown in Figs.(3, 4 & 5) which minimum levels at this case. With increasing the
represent the cases at N=5, 10 & 15, and as it is value of negative buoyancy ratio as it is shown in
clear from these figures, the solutal buoyancy Figs.[(7.a), (7.b) & (7.c)] which represent the
force which increases with the buoyancy ratio will flow case at N=-5, the negative solutal buoyancy
cooperate with the thermal buoyancy and they'll effect will be greater than the thermal buoyancy
drive the flow in the same direction to form a effect, and as a result of the total negative
cooperator flow, where the stream function levels buoyancy effect, the fluid at the hot wall will be
will increase with increasing in positive buoyancy heavier than it in other locations while the fluid at
ratio because of the increase of the total buoyancy the cold wall will be lighter than it in other
force due to both the thermal and solutal locations, so, the fluid particles move downwards
diffusing, while as it is shown from the contours along the hot wall while they move upwards along
of fluid phase temperature, its distribution will the cold wall, and thus, the flow will take the
keep on the same previous behavior, but the direction of counterclockwise. The fluid phase
gradients will be stronger than them at N=0 and temperature contour at N=-5 indicates that the
increase directly with increase of the buoyancy temperature levels will decrease gradually from
ratio, also a similar behavior to the fluid the hot wall towards the cold wall, where the hot
temperature distribution will appear in the fluid descends downwards along the hot wall and
contours of the solutal concentration, but their rises up along the cold wall because of the large
gradients will be relatively less than the gradients negative effect of solutal buoyancy. Finally the
of the fluid phase contours because of the absence solutal concentration contour at N=-5 indicates
of internal transference sources as it is clear from that the solutal concentration will be maximum at
eq.(23.e). Figs.(6, 7, 8, 9) the cases of negative the hot regions and decreases gradually with the
buoyancy ratio N<0, where the negative value of decrease of the fluid phase temperature. Figs.(8 &
buoyancy ratio means that the value of solutal 9) represent the flow case at N=-10 & -15
expansion coefficient βC is negative too, and as it respectively, and as it is clear that the levels of
is clear from the density definition in eq.(1), the counterclockwise stream function will increase

128
 Number 1 Volume 18 January 2012 Journal of Engineering

with the increase of negative value of buoyancy the value of local Nusselt number will be
ratio, while the fluid phase temperature and maximum at the top of the hot wall and it
solutal concentration keep the same previous descends gradually with dropping down to the
behaviours but their gradients increase with the wall bottom because the temperature gradient will
increase of negative value of buoyancy ratio. be very strong at the top and it decreases
Figs.(10.a & 10.b) indicate the variation of gradually to the bottom. Fig.(12) represents the
velocity components with buoyancy ratio at the variation of local Sherwood number with
intermediate vertical and horizontal locations buoyancy ratio (which is analog to Nusselt
-1
respectively for Da=10 and ε=0.9, where as it is number in heat transfer, where it gives an
shown in Fig.(10.a) the horizontal velocity indication to the rate of mass transfer through the
component at the positive values of buoyancy solution which happens by solutal concentration
ratio will direct to the right in top half of the difference), where it is clear that the distribution
enclosure while it will be in opposite direction in of local Sherwood number will be similar to the
the bottom half of the enclosure because of the distribution of local Nusselt number and for the
positive total buoyancy effect, while we'll note the same mentionable reasons previously. Fig.(13)
conversion in flow direction nearly at N=-1 where explains the variation of the mean Nusselt number
the horizontal velocity begins to reverse its with buoyancy ratio for different values of Darcy
direction as it happens when the value of negative number at ε=0.9, generally it is clear that the
buoyancy ratio becomes N<-1, also as it is shown minimum rate of heat transfer is at N=-1 because
in Fig.(10.b), the vertical velocity component at of the minimum temperature gradients at that
the positive values of buoyancy ratio will direct case, after that, heat transfer levels begin to
upwards in the left half of the enclosure while it increase with the increase of each one of positive
will be in opposite direction in the right half of the or negative buoyancy ratios, also it is clear that
enclosure because of the positive total buoyancy the values of Nusselt number will be lower than
effect, where it begins to reverse its direction at them at the same values of positive buoyancy ratio
the negative buoyancy ratios of N<-1. Fig.(11) because the thermal and solutal buoyancy effects
represents the variation of local Nusselt number at negative buoyancy ratios will be in opposite
-1
with buoyancy ratio at Da=10 and ε=0.9, where direction while they in same direction at positive
it is clear that the value of local Nusselt number at buoyancy ratios to form (assisting flow), and
N≥0 will be maximum at the bottom of the left- finally it was noted that the mean Nusselt number
hand vertical wall and it descends gradually with generally increases with Darcy number increase
rising to the wall top because the temperature because of the decreasing in overall bed resistance
gradient will be very strong at the bottom and it to the flow inside it as a result to the increasing in
decreases gradually to the top, also it is clear the the volume of passable paths of fluid through the
increasing in the local Nusselt number levels with packed bed. Fig.(14) explains the variation of the
the increase of positive buoyancy ratio, while the mean Nusselt number with buoyancy ratio for
minimum levels of it will be at N=-1 because the different values of porosity at Da=10-1, where it
temperature gradients were minimum at that was noted that when porosity equals to a value in
buoyancy ratio value, also at N<-1 it is clear that the range of (ε=0.5~0.8), the values of mean

129
Ahmed N. Mehdy DOUBLE DIFFUSIVE FREE CONVECTION IN A PACKED
BED SQUARE ENCLOSURE BY USING LOCAL THERMAL
NON-EQUILIBRIUM (LTNE) MODEL
Nusselt number will increase gradually with the similarity in Nusselt number behavior with the
increase of porosity in that range, but when the mentionable study, but there is a small difference
porosity increases above that range, the values of between these values because of the using of
mean Nusselt number will decrease, and for (LTNE) model in the present study which causes
explaining this phenomenon clearly we must more heat transfer due to both the fluid and solid
return to Eq.(20.a), where it is noted that the phases.
whole value of local Nusselt number is
multiplicand by the porosity while the second CONCLUSIONS
term in this equation (solid phase term) was This paper has presented a numerical
multiplicand by the inverse of dimensionless investigation of double-diffusive flow in a packed
thermal conductivity Λ-1 which decreases with the bed square enclosure by using local thermal non-
[ ]
increase of porosity Λ−1 = (1 − ε ) k s ε k f , so, equilibrium (LTNE) model, and from the obtained
results, the following conclusions are drawn;
when the porosity values equal to or less than
1- At the positive buoyancy ratios N≥0, the flow
(0.8), the value of Nusselt number will increase,
takes the clock wise direction, while at N=-1
while it decreases after that value of porosity
because the value of dimensionless thermal the flow begins to reverse its direction as it

conductivity will be very small and it causes happens at N<-1.

decreasing in Nusselt number more than the 2- Generally, levels of flow, heat transfer and
increasing in it because of porosity. Fig.(15) mass transfer increase with the increase of
shows the variation of the mean Sherwood both the positive or negative buoyancy ratio
number with buoyancy ratio for different values as a result of increasing in the total buoyancy
of Darcy number at ε=0.9, where Sherwood effects.
number will comport a similar behavior to Nusselt 3- The values of Nusselt number increase with
number variation with buoyancy ratio and Darcy increase of Darcy Number.
number and for the same mentionable reason 4- The values of Nusselt number increase with
previously, while Fig.(16) explains the variation increase of porosity until the porosity reaches
of the mean Sherwood number with buoyancy
a certain value of about (ε ≈ 0.8) where the
ratio for different values of porosity at Da=10-1,
value of mean Nusselt number will decrease
where it is clear that the values of Sherwood
after that value of porosity.
number will increase with increase of packed bed
5- The values of Sherwood number increase
porosity because the flow levels will increase as a
with increase of both Darcy Number and the
result of decreasing in the occupied volume by the
porosity.
bed through enclosure, where the overall packed
REFERENCES
bed resistance to the flow will decrease. To
exhibit the reliability of the presented results, the A. Amiri and K. Vafai, Transient analysis

variation of mean Nusselt number with the of incompressible flow through a packed bed,

buoyancy ratio at GrT≈1.4×105 was compared International Journal of Heat and Mass Transfer,
with results of (Rahli and Bouhadef, 2004) as it 41, 4259-4279, 1998.
is shown in Fig.(17), where it is clear the

130
 Number 1 Volume 18 January 2012 Journal of Engineering

A. K. Bukhari, Double diffusive International Journal of Heat and Mass Transfer

convection in a horizontal porous layer 50, 3684-3689, 2007.
superposed by a fluid layer, Umm Al-Qura Univ. R. A. Mohamed, Double-diffusive
J. Sci. Med. Eng. Vol. 15, No. 2, pp. 95-113, convection-radiation interaction on unsteady
2003. MHD flow over a vertical moving porous plate
I. Hajri*, A. Omri and S. Ben Nasrallah, with heat generation and soret effects, Applied
A numerical model for the simulation of double- Mathematical Sciences, Vol. 3, no. 13, 629 - 651,
diffusive natural convection in a triangular cavity 2009.
using equal order and control volume based on the R. C. Chaudhary and Arpita Jain,
finite elements method, Desalination 206, 579- Combined heat and mass transfer effects on MHD
588, 2007. free convection flow past an oscillating plate
K. Khanafer and K. Vafai, Double- embedded in porous medium, Rom. Journ. Phys.,
diffusive mixed convection in a lid-driven Vol. 52, Nos. 5–7, P. 505–524, Bucharest, 2007.
enclosure filled with a fluid saturated porous S.C. Saha, M.A. Hossain, Natural
medium, Numerical Heat Transfer, Part A, 42: convection flow with combined buoyancy effects
465-486, 2002. due to thermal and mass diffusion in a thermally
M. Mamou, P. Vasseur and M. Hasnaoui, stratified media, Nonlinear Analysis: Modelling
On numerical stability analysis of double- and Control, Vol. 9, No. 1, 89–102, 2004.
diffusive convection in confined enclosures, J. Versteeg H. K. and Malalasekera W., An
Fluid Mech., vol. 433, pp. 209-250, 2001. introduction to computational fluid dynamics the
Md. Shariful Alam, M. M. Rahman and Md. finite volumes method, Longman Group Ltd,
Abdul Maleque, Local similarity solutions for 1995.
unsteady MHD free convection and mass transfer Y. Masuda, M. Yonaya, T. Ikeshoji, S.
flow past an impulsively started vertical porous Kimura, F. Alavyoon, T. Tsukada, M. Hozawa,
plate with Dufour and Soret effects, Thammasat Oscillatory double-diffusive convection in a
Int. J. Sc. Tech., Vol. 10, No. 3, July-Sept., 2005. porous enclosure due to opposing heat and mass
O. Rahli and K. Bouhadef *, Double- fluxes on the vertical walls, International Journal
diffusive natural convection in a partially porous of Heat and Mass Transfer 45, 1365-1369, 2002.
square enclosure; effect of the inclination, Yoshio Masuda, Michio Yoneya, Akira
Laboratoire LTPMP, Faculté de Génie Mécanique Suzuki, Shigeo Kimura and Farid Alavyoon,
et Génie des Procédés, USTHB, Alger – Algérie, Numerical analysis of double-diffusive convection
2004. in a porous enclosure due to opposing heat and
Q. W. Wang*, M. Zeng, Z. P. Huang, G. mass fluxes on the vertical walls -Why does
Wang, H. Ozoe, Numerical investigation of peculiar oscillation occur?-, Research Center for
natural convection in an inclined enclosure filled Compact chemical Process, National Institute of
with porous medium under magnetic field, Advanced Industrial Science and Technology, 4-

131
Ahmed N. Mehdy DOUBLE DIFFUSIVE FREE CONVECTION IN A PACKED
BED SQUARE ENCLOSURE BY USING LOCAL THERMAL
NON-EQUILIBRIUM (LTNE) MODEL
2-1 Nigatake, Miyagino-ku, Sendai 983-8551,
Japan, 2008.
Nomenclature Greek symbols
a side length of the enclosure (m) α thermal diffusivity (m2 s-1)
asf specific surface area of the packed bed (m-1) µ dynamic viscosity (kg m-1 s-1)
C dimensionless solutal concentration ν kinematic viscosity (m2 s-1)
c solutal concentration (kmol m-3) ρ density (kg m-3)
cp specific heat at constant pressure (J kg-1 K-1) βT thermal expansion coefficient (K-1)
D concentration diffusion coefficient (m2 s-1) βC solutal expansion coefficient (kmol-1 m3)
Da Darcy number θ dimensionless temperature
dp sphere particle diameter (m) ψ dimensionless stream function
F geometric function ω dimensionless vorticity
GrC solutal Grashof number Φ general scalar dependent variable
GrT Grashof number Γ diffusion coeffecient
g gravitational acceleration (m s-2) ε porosity
hsf solid-fluid heat transfer coefficient(Wm-2K-1) Λ dimensionless thermal conductivity
k thermal conductivity (W m-1 K-1) ξ dimensionless solid-to-fluid heat transfer
K permeability (m2) coefficient
N buoyancy ratio
Nu Nusselt number Subscripts
p pressure (Pa) f fluid
P dimensionless pressure feff effective properties for fluid
Pr Prandtl number H high
Sc Schmidt number i axis indication
Sh Sherwood number L low
T temperature (K°) l local value
u, v velocity components (m s-1) m mean value
U, V dimensionless velocity components s solid
x, y x-, y-coordinates (m) seff effective properties for solid
X, Y dimensionless coordinates

Packed Bed

TH TL
a
cH g cL
y

Fig.(1) schematic diagram of the physical system

132
 Number 1 Volume 18 January 2012 Journal of Engineering

1 1 1

0.9 0.9 0.9

0.9375 0.9375
0.8 -0.001 0.8 0.875 0.8 0.875
-0.009 0.8125 0.8125
0.7 -0.02 0.7 0.75 0.7 0.75
-0.035 0.6875 0.6875
0.6 -0.05 0.6 0.625 0.6 0.625
-0.065 0.5625 0.5625
0.5 -0.08
Y

0.5 0.5 0.5

Y
0.5

Y
-0.095 0.4375 0.4375
0.4 -0.11 0.4 0.375 0.4 0.375
-0.12 0.3125 0.3125
0.3 -0.13 0.25 0.3 0.25
0.3
-0.137 0.1875 0.1875
-0.142 0.125 0.125
0.2 0.2 0.2
0.0625 0.0625
0.1 0.1 0.1

0 0 0
0 0.25 0.5 0.75 1 0 0.25 0.5 0.75 1 0 0.25 0.5 0.75 1
a) Streamlines
X b) Fluid phaseX temperature c) Solutal concentration
X
-1
Fig. (2) Properties distribution at; N=0, Da=1×10 & ε =0.9
1 1 1
0.9 0.9 0.9
0.9375 0.9375
0.8 -0.001 0.8 0.875 0.8 0.875
-0.03 0.8125
0.7 -0.1 0.8125
0.7 0.75 0.7
-0.17 0.75
0.6875 0.6875
0.6 -0.24 0.6 0.625 0.6 0.625
-0.3 0.5625
-0.35 0.5625
0.5
Y

0.5 0.5
Y

0.5 0.5

Y
-0.4 0.4375
-0.44 0.4375
0.4 0.4 0.375 0.375
-0.47 0.4
0.3125 0.3125
0.3 -0.49 0.25
-0.51 0.3 0.3 0.25
0.1875 0.1875
-0.52 0.125
0.2 0.2 0.125
0.0625 0.2
0.0625
0.1 0.1 0.1
0 0
0 0.25 0.5 0.75 1 0
0 0.25 0.5 0.75 1 0 0.25 0.5 0.75 1
a) Streamlines
X
b) Fluid phaseX temperature X
c) Solutal concentration
Fig. (3) Properties distribution at; N=5, Da=1×10-1 & ε =0.9

1 1 1

0.9 0.9 0.9

0.9375 0.9375
0.8 -0.001 0.8 0.875 0.8 0.875
-0.03 0.8125 0.8125
0.7 -0.1 0.7 0.75 0.7 0.75
-0.18 0.6875 0.6875
0.6 -0.25 0.6 0.625 0.6 0.625
-0.35 0.5625 0.5625
0.5 -0.44
Y

0.5 0.5 0.5 0.5

Y

-0.5 0.4375 0.4375

0.4 -0.55 0.4 0.375 0.4 0.375
-0.59 0.3125 0.3125
0.3 -0.62 0.3 0.25 0.3 0.25
-0.65 0.1875 0.1875
-0.67 0.125 0.125
0.2 0.2 0.2
0.0625 0.0625
0.1 0.1 0.1

0 0 0
0 0.25 0.5 0.75 1 0 0.25 0.5 0.75 1 0 0.25 0.5 0.75 1
a) Streamlines
X b) Fluid phaseX temperature c) Solutal concentration
X

Fig. (4) Properties distribution at; N=10, Da=1×10-1 & ε =0.9

1 1 1

0.9 0.9 0.9

-0.001 0.9375 0.9375
0.8 -0.03 0.8 0.875 0.8 0.875
-0.1 0.8125 0.8125
0.7 -0.18 0.7 0.75 0.7 0.75
-0.25 0.6875 0.6875
0.6 -0.35 0.6 0.625 0.6 0.625
-0.44 0.5625 0.5625
0.5 -0.52
Y

0.5 0.5 0.5 0.5

Y

-0.58 0.4375 0.4375

0.4 -0.63 0.4 0.375 0.4 0.375
-0.67 0.3125 0.3125
0.3 -0.7 0.3 0.25 0.3 0.25
-0.73 0.1875 0.1875
0.2 -0.75 0.125 0.125
0.2 0.2
-0.77 0.0625 0.0625
0.1 0.1 0.1

0 0 0
0 0.25 0.5 0.75 1 0 0.25 0.5 0.75 1 0 0.25 0.5 0.75 1
X X X

a) Streamlines b) Fluid phase temperature c) Solutal concentration

-1
Fig. (5) Properties distribution at; N=15, Da=1×10 & ε =0.9
133
 Ahmed N. Mehdy DOUBLE DIFFUSIVE FREE CONVECTION IN A PACKED
BED SQUARE ENCLOSURE BY USING LOCAL THERMAL
NON-EQUILIBRIUM (LTNE) MODEL
1 1 1

0.9 0.9 0.9

0.08
0.07 0.9375 0.9375
0.8 0.06 0.8 0.875 0.8 0.875
0.05 0.8125 0.8125
0.7 0.04 0.7 0.75 0.7 0.75
0.03 0.6875 0.6875
0.6 0.02 0.6 0.625 0.6 0.625
0.01 0.5625 0.5625
0.5
Y

0.0025 0.5 0.5 0.5 0.5

Y
0.001 0.4375 0.4375
0.4 -0.001 0.4 0.375 0.4 0.375
-0.005 0.3125 0.3125
0.3 -0.01 0.3 0.25 0.3 0.25
-0.013 0.1875 0.1875
0.2 -0.016 0.2 0.125 0.2 0.125
-0.02 0.0625 0.0625
0.1 -0.023
0.1 0.1

0 0 0
0 0.25 0.5 0.75 1 0 0.25 0.5 0.75 1 0 0.25 0.5 0.75 1
X a) Streamlines
b) Fluid phaseXtemperature c) Solutal concentration
X
-1
Fig. (6) Properties distribution at; N=-1, Da=1×10 & ε =0.9
1 1 1

0.9 0.9 0.9

0.515 0.9375 0.9375
0.8 0.8 0.875 0.8 0.875
0.505
0.49 0.8125 0.8125
0.7 0.7 0.75 0.7 0.75
0.47
0.44 0.6875 0.6875
0.6 0.4 0.6 0.625 0.6 0.625
0.35 0.5625 0.5625
0.5
Y

0.5 0.5 0.5 0.5

Y

Y
0.3
0.25 0.4375 0.4375
0.4 0.2 0.4 0.375 0.4 0.375
0.15 0.3125 0.3125
0.3 0.1 0.3 0.25 0.3 0.25
0.05 0.1875 0.1875
0.2 0.01 0.2 0.125 0.2 0.125
0.0625 0.0625
0.1 0.1 0.1

0 0 0
0 0.25 0.5 0.75 1 0 0.25 0.5 0.75 1 0 0.25 0.5 0.75 1
a) Streamlines
X b) Fluid phaseX temperature c) Solutal concentration
X
-1
Fig. (7) Properties distribution at; N=-5, Da=1×10 & ε =0.9
1 1 1

0.9 0.9 0.9

0.67
0.9375 0.9375
0.8 0.655 0.8 0.8
0.875 0.875
0.63
0.8125 0.8125
0.7 0.6 0.7 0.7
0.75 0.75
0.56
0.6875 0.6875
0.6 0.51
0.6 0.625 0.6 0.625
0.46
0.5625 0.5625
0.4
0.5
Y

0.5 0.5 0.5 0.5

Y

0.35
0.4375 0.4375
0.3
0.4 0.4 0.375 0.4 0.375
0.25
0.3125 0.3125
0.2
0.3 0.3 0.25 0.3 0.25
0.15
0.1875 0.1875
0.1
0.2 0.2 0.125 0.2 0.125
0.05
0.0625 0.0625
0.01
0.1 0.1 0.1

0 0 0
0 0.25 0.5 0.75 1 0 0.25 0.5 0.75 1 0 0.25 0.5 0.75 1
a) Streamlines
X b) Fluid phaseX temperature c) Solutal concentration
X
-1
Fig. (8) Properties distribution at; N=-10, Da=1×10 & ε =0.9

1 1 1

0.9 0.9 0.9

0.77
0.9375 0.9375
0.8 0.755 0.8 0.8
0.875 0.875
0.73
0.8125 0.8125
0.7 0.7 0.7 0.7
0.75 0.75
0.67
0.6875 0.6875
0.6 0.62 0.6 0.6
0.625 0.625
0.57
0.5625 0.5625
0.5
0.5
Y

0.5 0.5 0.5 0.5

Y
Y

0.42
0.4375 0.4375
0.33
0.4 0.4 0.375 0.4 0.375
0.25
0.3125 0.3125
0.17
0.3 0.3 0.25 0.3 0.25
0.1
0.1875 0.1875
0.05
0.2 0.2 0.125 0.2 0.125
0.01
0.0625 0.0625
0.1 0.1 0.1

0 0 0
0 0.25 0.5 0.75 1 0 0.25 0.5 0.75 1 0 0.25 0.5 0.75 1
a) Streamlines
X b) Fluid phaseX temperature c) Solutal concentration
X

Fig. (9) Properties distribution at; N=-15, Da=1×10-1 & ε =0.9

134
 Number 1 Volume 18 January 2012 Journal of Engineering

1
10 N=-15
0.9 N=-10
7.5 N=-5
0.8 N=-1
N=0
5 N=5
0.7
N=-15 N=10
N=-10 2.5 N=15
0.6
N=-5
N=-1

V
0.5 0
Y

N=0
N=5
0.4 -2.5
N=10
N=15
0.3
-5
0.2
-7.5
0.1
-10
0
-3 -2.5 -2 -1.5 -1 -0.5 0 0.5 1 1.5 2 2.5 3 0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1
a) U- velocity component at the b) V- velocity component at the
U X
intermediate vertical location intermediate horizontal location
Fig. (10) Variation of velocity components with buoyancy ratio
at the intermediate vertical and horizontal locations for Da=1×10-1 & ε =0.9
1 1

0.9 0.9

0.8 0.8

0.7 N=-15 0.7 N=-15

N=-10 N=-10
0.6 N=-5 0.6 N=-5
N=-1 N=-1
0.5 0.5
Y

N=0 N=0
N=5 N=5
0.4 N=10 0.4 N=10
N=15 N=15
0.3 0.3

0.2 0.2

0.1 0.1

0 0
0 5 10 15 20 25 30 35 40 45 50 55 0 2 4 6 8 10 12 14 16 18 20 22 24 26
Fig. (11) Variation of local Nusselt number Fig. (12) Variation of local Sherwood number
Nu l Sh l
with buoyancy ratio at ε=0.9 & Da=1×10-1 with buoyancy ratio at ε=0.9 & Da=1×10-1
22 22
22
Da=1.E-5
20 Da=1.E-5 20
20 Da=1.E-4
Da=1.E-4
Da=1.E-3
Da=1.E-3
18 Da=1.E-2 18
18 Da=1.E-2
Da=1.E-1
Da=1.E-1
16
16 16

14
14 14
Nu m
Numm

12
12 12
Nu

10
10 10

88 8

66 6 Porosity=0.5
Porosity=0.6
Porosity=0.7
44 4 Porosity=0.8
Porosity=0.9
22
2
-15
-15 Fig.
-10 (13) Variation
-10 -5
-5 00 of mean
55 Nusselt
10
10 15
15
-15
Fig.
-10
(14) Variation
-5 0
of mean
5
Nusselt
10 15

number with Darcy N

N number at ε=0.9 N
number with porosity at Da=1.E-1
135
Ahmed N. Mehdy DOUBLE DIFFUSIVE FREE CONVECTION IN A PACKED
BED SQUARE ENCLOSURE BY USING LOCAL THERMAL
NON-EQUILIBRIUM (LTNE) MODEL

10 10
Da=1.E-5
9 Da=1.E-4 9
Da=1.E-3
Da=1.E-2
8 Da=1.E-1 8

7 7

6 6

Sh m
Sh m

5 5

4 4

3 3 Porosity=0.5
Porosity=0.6
Porosity=0.7
2 2 Porosity=0.8
Porosity=0.9

1 1
-15 -10 -5 0 5 10 15 -15 -10 -5 0 5 10 15
Fig. (15) Variation of
N mean Sherwood Fig. (16) Variation of
N mean Sherwood
number with Darcy number at ε=0.9 number with porosity at Da=1.E-1

11
Rahli' s Work at; Da=1.E-5
Present Work at; Da=1.E-5
10
Rahli' s Work at; Da=1.E-4
Present Work at; Da=1.E-4
9

7
Nu m

1
-10 -5 0 5 10
5
Fig. (17) Comparison the variation of mean Nusselt
N number with buoyancy ratio at GrT≈1.4×10
with the work of (Rahli and Bouhadef, (2004)

136